The highly specific and complex nature of olfactory sensing systems has inspired researchers to develop vapor phase chemical detection systems, commonly referred to as “electronic noses.” These devices and systems are currently being used in biotechnology, as well as applications in medicine, the environment, the food industry, and most recently law enforcement applications. Interdiction efforts continue in the search for technologies which can provide an inexpensive alternative to dogs as detectors of narcotics and explosives. One of the principal motivations for the development of electronic noses for such applications is the expense associated with the handling, training, and care of such dogs, Furthermore, it is often unclear what chemicals the dogs are actually detecting, and response vary considerably amongst dogs. As an example, not all dogs do respond to the same cocaine sample. Although the dogs have proved to be a highly useful tool in detecting illicit materials, they have inherent limitations in reliability, as one would never accept data from an instrument without having a solid idea of the physical mechanism behind a detection event.
Acoustic sensors represent a long-standing an approach for high-precision sensing. Quartz crystal microbalances (QCMs) have been utilized since the 1950s to monitor the thickness of metals being deposited on wafers in evaporation systems. Such sensors have leveraged investments in other technologies, namely, oscillator designs and electronic frequency counters. In addition, the investments in frequency control and radar during World War II and for the quarter century that followed led to a detailed understanding of the temperature characteristics of quartz, largely at the Ft. Monmouth, N.J., Army Research Lab in the United States. In cuts of quartz such as the AT-cut used for QCM, the linear expansion of the material with increasing temperature is compensated by an increase in the acoustic velocity such that the round trip delay for an acoustic wave in a resonator does not vary with temperature. However, current implementations of QCMs, as used to detect chemicals in the liquid phase, suffer from the inability to distinguish molecular detection events from noise, such as the binding of extraneous substances to the detection device. Furthermore, existing models of recognition events are static, and do not account for time variations in the state of detection devices, and thereby ignore important distinguishing characteristics of detection events. One such static model is the Sauerbrey equation, i.e.,:
      Δ    ⁢                  ⁢    •    ⁢                  ⁢    f    =      -                  2        ⁢                                  ⁢                  f          o          2                ⁢                                  ⁢                  ρ          s                                      V          a                ⁢                  ρ          r                    where Δf is the resonator frequency shift; fo is the resonator center frequency; ρs is the mass density per m2 of analyte attached to the surface; Vα is the acoustic wave velocity in the resonator, and ρr, is the volume mass density of resonator material. (further described in “Use of quartz vibrator for weighing thin films on a microbalance,” Z Phys., vol. 155, pp. 206-210, 1959.). By failing to account for changes in response over time, such models are often inadequate to distinguish molecular recognition events from noise.
Acoustic wave “biosensors” are distinguished from chemical sensors in that they use a molecule of biological origin (e.g., antibody, cell, enzyme, protein) immobilized onto a surface as the chemically sensitive film on a device. In the prior art, the detection of the presence of entities of biological origin, such as proteins or cells, has taken place in liquids as a requirement. Accordingly, there is a need for an acoustic sensor and models for signature recognition which have one or more of the following features: (1) allow for the detection of targets other than biomolecules, (2) do not limit the detection environment to the liquid phase; (3) allow detection of bio-molecules in the vapor phase; (4) allow for the detection of molecular recognition events through time-dependent signatures, to improve accuracy and speed of detection.